ABSTRACT Background. Veterans with the severe mental illnesses (SMI) of schizophrenia and bipolar disorder, as well as those with post-traumatic stress disorder and major depression, undergo surgery at a higher rate than veterans without SMI (2.5% vs 1.4% in fiscal 2005). These mental illnesses are common in the VA and have a devastating impact on all aspects of health. Yet our systematic literature review (1966-2007) identified only 12 studies regarding perioperative clinical outcomes for SMI patients. This meager evidence base highlights an important gap in knowledge regarding surgical outcomes of patients with SMI. Most surgical quality of care studies derive from the VA's National Surgical Quality Improvement Program (NSQIP), however, NSQIP collects no information about preoperative psychiatric diagnoses. Therefore, NSQIP cannot even begin to explore perioperative risks and outcomes for SMI patients. SMI patients may fare poorly after surgery because of later presentation of surgical disease, lifestyle factors, and multiple chronic medical conditions. Therefore we plan to measure rates of surgery and to evaluate surgery outcomes among VA patients with and without SMI. Objectives. (1) Compare surgery rates among VA patients by SMI status (schizophrenia, bipolar disorder, PTSD, MDD, no mental illness). (2) Assess 30-day, 90-day, and 1-year postoperative mortality by SMI status. (3) Assess 30-day, 90-day, and 1-year postoperative complications (e.g., heart attack, venous thromboembolism, ICU admission, pneumonia, respiratory failure, sepsis, wound infection) by SMI status. Methods. Our retrospective cohort study will use archival VA data to compare outcomes of surgery among patients with schizophrenia, bipolar disorder, post-traumatic stress disorder, or major depressive disorder to outcomes of surgery patients without prior diagnosis of any mental illness over the 5-year period FY05-FY09. We will calculate rates of surgery, postoperative mortality, and postoperative complications, contrasting VA patients with and without SMI. Multivariable regression analyses will assess the impact of SMI on surgery outcomes, after adjusting for demographic and clinical covariates. Exploratory analyses will assess potential differences in complexity of surgical disease by SMI status using Relative Value Units (RVUs) for selected procedures, and the impact of organizational characteristics on postoperative outcomes. Impacts. This study will produce new knowledge in the area of surgical treatment and severe mental illness. The project will produce previously unexamined data on surgical care for the population of veterans with severe mental illness, a patient subgroup with complex healthcare needs. The proposed work will lay the groundwork for identifying and developing interventions to improve perioperative care of these patients.